Labor of Love: Immigrant–Native Intermarriage and Labor Force Outcomes Across European Union Member States Open Access

Abstract

Immigrant–native intermarriage has been shown to improve immigrant labor force outcomes. A parallel line of research demonstrates that immigrant employment outcomes are strongly influenced by the sociopolitical climate and overall reception immigrants receive upon arrival in the host society. Our research spans both strands of literature to address unresolved questions about how the association between immigrant–native intermarriage and foreign-born occupational status changes across host societies on the basis of their national-level polices and views toward outsiders. To assess this, we analyze substantive changes in immigrant occupational attainment across twenty European Union countries from 2008 to 2018 using individual-level, cross-sectional data from the European Union Labor Force Survey on occupational attainment merged with country-level data from the European Social Survey on attitudes toward outsiders and the Migrant Integration Policy Index on integration policies. Our findings show that the association between intermarriage and immigrant occupation status is stronger in countries with more welcoming policies, whereas attitudes toward outsiders do not have a significant effect net of the moderating role of policy. The novel use of occupational status along with the moderating effect of policy provides further evidence that closed societies impede immigrant integration while deepening immigrant marginalization.

Introduction

Measures of integration¹ have long been of central importance in migration studies (Davis 1941; Gordon 1964; Song 2009). This focus makes sense because factors that impede or accelerate the process of incorporation impact the well-being of group members far beyond the first generation of arrivals (Gorodzeisky and Semyonov 2017; Meurs et al. 2006; Suárez-Orozco et al. 2015; Verhaeghe et al. 2013). Variables affecting immigrant integration can be broadly separated into two groups: those at the societal level and those at the individual level.

After laws and policies regulating the admission and settlement of foreign nationals, the most influential societal-level factor for immigrant integration is the social climate toward outsiders within the receiving society (Alba and Foner 2015; Fussell 2014; Reitz 1998). Scholars have operationalized social climate using a variety of survey data assessing attitudes or beliefs held and actions taken toward immigrants or ethnic groups in general (Ceobanu and Escandell 2010; Mayda 2006; Rustenbach 2010) or toward specific groups (Hellwig and Sinno 2017; Schachter 2016). Among individual-level factors, such as age (Schaafsma and Sweetman 2001), gender (Kesler 2018; Korteweg 2017), race and ethnicity (Favell 2022), language ability (Harder et al. 2018), educational credentials (Kogan 2016), and the presence of minor children in the household (Chiswick and Houseworth 2011), immigrant–native intermarriage² is of special concern as it is conceptualized as both a factor hastening immigrant integration (Basu 2017) and an indicator of the degree of integration or assimilation³ (Gordon 1964) of an immigrant group within receiving societies (Bean and Stevens 2003; Song 2009; Stevens et al. 2012). Several contemporary analyses have demonstrated compelling evidence that immigrant–native intermarriage plays an important role in transmitting cultural and economic resources, as well as in improving labor market outcomes (Dribe and Nystedt 2014; Furtado and Song 2015; Stevens et al. 2012).

There is growing consensus within migration studies that immigrant–native intermarriage may help immigrants access parts of society that are effectively closed to nonmajority members, especially within the labor market (Dribe and Nystedt 2014; Furtado and Song 2015; Meng and Gregory 2005). It is still unclear whether the benefit of intermarriage changes as societies become more open or closed to outsiders over time (Ivory 2023). This question is all the more important given spikes in nativist and racist sentiment during periods of perceived economic precarity in the early twenty-first century across much of Europe and North America (Bleich 2011; Fekete 2018; Hooghe and Dassonneville 2018; Kuntz et al. 2017; O'Brien and Eger 2021; Taras 2012) and the wide range of gender inequality experienced in the labor market across the same geographic area (Adserà and Chiswick 2007; Kesler 2018).

This article asks the following questions: Does the positive association between immigrant–native intermarriage and an immigrant's occupational status vary across host societies with more or less welcoming policies or views toward outsiders? And if so, by what degree? We analyze individual-level, cross-sectional data from the European Union Labor Force Survey (EU-LFS) augmented with country-level indicators from the European Social Survey (ESS) and Migrant Integration Policy Index (MIPEX) to assess substantive changes in immigrant labor force outcomes across twenty European Union (EU) countries from 2008 to 2018, and we link these changes to aspects of the institutional context. Consistent with the literature on immigrant–native intermarriage for other labor market outcomes, we find that intermarriage is positively associated with immigrants’ occupational status for both genders despite foreign-born women's greater exposure to gender inequality in receiving society labor markets (Ivory et al. 2024; Kesler 2018) and lower likelihood of representation in the formal economy (Phizacklea 2022). We extend prior findings by showing that this association is stronger in host societies with more welcoming integration policies. Conversely, this association is not significantly moderated by the societal attitude toward immigrants. Thus, we contribute to the literature on the labor market effects of immigrant–native intermarriage by drawing attention to the context of receiving societies and further refining mechanisms to quantify the magnitude of the association between intermarriage and foreign-born labor force outcomes, particularly with the novel use of occupational status instead of measures of labor force participation or wages.

Background

Intermarriage and Its Labor Market Effects

As previously mentioned, immigrant–native intermarriage is central in sociological inquiries into immigrant integration. Whereas earlier studies (Davis 1941; Gordon 1964; Warner and Srole 1945) were primarily concerned with the role of intermarriage in social integration, contemporary studies now regularly assess how it moderates the interaction of various social and economic measures (Lichter et al. 2015; Meng and Gregory 2005; Stevens et al. 2012). Commonly used economic indicators focus on labor market outcomes, such as labor force participation (Basu 2017), income (Meng and Gregory 2005), employment/unemployment/underemployment (Furtado and Theodoropoulos 2010), working hours (Nottmeyer 2011), and occupational prestige or status (Rodríguez-García et al. 2015).

Although the relationship between immigrant–native intermarriage and foreign-born workers’ labor market outcomes has been well researched (Basu 2015; Bean and Stevens 2003; Furtado and Song 2015; Furtado and Theodoropoulos 2010; Meng and Gregory 2005), there is no consensus regarding the direction of the association—which may be positive or negative depending on societal expectations around gender and family (Ivory et al. 2024)—or which indicators provide the most useful information (Furtado and Theodoropoulos 2010).

The nexus between immigrant–native intermarriage and gendered outcomes in the labor market has also produced several important, if sometimes conflicting, insights. Although all foreign-born partners gain some socioeconomic advantage by marrying a native, intermarriage tends to confer more immediate labor market benefits to foreign-born men than women (Chiswick and Houseworth 2011; Furtado 2012). Popular explanations for this inequality relate to the greater importance of a foreign-born spouse in protecting foreign-born men from difficult labor market conditions. Foreign-born men are more likely to be the target of explicit exclusion or discrimination in European labor markets, while foreign-born women are more likely to be covered by gender-inclusive employment protections (Platt et al. 2022). In addition, the wage gap between foreign- and native-born women is less substantial than that between foreign- and native-born men (Cantalini et al. 2023). Some accounts, such as Rodríguez-García et al. (2015) and Basu (2017), provided evidence that certain groups of foreign-born women receive a larger labor market boost in the short term than their foreign-born male counterparts. Moderating factors, such as ethnic origin (Guetto and Azzolini 2015) and religion (Lucassen and Laarman 2009), play a significant role in determining the differential benefit of intermarriage for foreign-born women and men but are not always enumerated.

Interestingly enough, few cross-national studies have assessed the interaction between occupational status and immigrant–native intermarriage. This is somewhat puzzling since the most widely used measure of occupational status, the International Socio-Economic Index of Occupational Status (ISEI), is more stable across both time and geographic region than income, working hours, labor force participation, and employment/unemployment (Ganzeboom et al. 1992; Ganzeboom and Treiman 2003). Occupational status is also uniquely suited to provide insights on the utility of intermarriage for foreign-born workers because it has been widely used within survey research to demonstrate the transmission of advantage or disadvantage across generations (Torche 2015) and even within native-born married couples (Shauman 2010). Prior research has assessed occupational status within interethnic and interracial unions, particularly those between Asian/Asian American women and their partners within the United States (Chen and Takeuchi 2011), finding that intermarried couples are far more likely to have similar occupational statuses than intramarried couples (Platt et al. 2022); moreover, where there are differences, the foreign-born partner is more likely to enter a union with higher occupational status (Guetto and Azzolini 2015). To our knowledge, no previous cross-national comparative studies have used occupational status as the main dependent variable in assessing the impact of immigrant–native intermarriage on labor market outcomes.

The Role of Immigrant Receiving Contexts on Labor Market Incorporation

Research has identified a variety of elements accelerating or impeding the participation and deeper penetration of immigrants in domestic labor markets (Kogan 2016). Moreover, these factors have been assessed at both the country and municipal levels across several European contexts. Immigration policy looms large because government rules and regulations often directly exclude some or all immigrants from accessing parts of the labor market or segregate immigrants into specific sectors. The connection between immigration policy and foreign-born employment outcomes has a long history in Western societies (Bloom et al. 1994; Mott 1966; North and Martin 1980; Parai 1975) and has been reaffirmed by contemporary investigations (Akbari and MacDonald 2014; Careja 2019; Cobb-Clark 2004). Several analyses demonstrated that foreign-born individuals residing in countries with recently adopted laws or policies restricting the legal admission of immigrants typically experience lower levels of employment (Akbari and MacDonald 2014; Careja 2019), longer spells of unemployment (Büchel and Frick 2005), and higher levels of segregation into poorly remunerated manual labor jobs (Dustmann and Frattini 2011) than foreign-born individuals residing in countries with more relatively open immigration policies.

Closely related to immigration policy is integration policy, which are laws and policies regulating the settlement and well-being of immigrants and their family members. Although not all countries codify integration policy at the national level, those that do tend to have higher levels of foreign-born employment and greater levels of skills matching than those that do not (Büchel and Frick 2005). While not specifically aimed at immigrants, social welfare policies have a positive correlation with better labor market outcomes for foreign-born workers in instances where such benefits are universally granted (Römer 2017), largely because a vibrant social safety net improves the health and well-being of all residents and is associated with positive spillover effects in the labor market (Römer 2017). Policies that promote gender equality (van Tubergen et al. 2004), provide sufficient support for family care (Drange and Telle 2015), and ensure enough monetary resources to take care of essential needs (Kesler 2015) are all associated with greater levels of foreign-born labor market participation. Lastly, a variety of negative native-born attitudes and beliefs about immigrants or various minority/outsider groups are associated with diminished foreign-born labor force outcomes. Countries with low levels of trust (van der Linden et al. 2017), high levels of animosity toward minority groups or immigrants (Alba and Foner 2015), and more widespread beliefs that immigrants take jobs that “belong” to native-born citizens (Dancygier and Laitin 2014) are more likely to have lower foreign-born labor force participation and greater levels of immigrant occupational segregation.

The relationship between the context of reception and immigrant occupational attainment has received heightened scholarly attention in recent years. The bulk of this research has emphasized the utility of interacting occupational attainment with labor market outcomes, such as labor force participation or wages, because some polices will affect one indicator but not the other (Platt et al. 2022). Some of the most important findings from this line of inquiry include linking policies hastening or impeding legal status acquisition to labor market outcomes (Corrigan 2015); demonstrating that the link between immigration policies and labor market outcomes is most impactful for lower skilled immigrants (Helbling et al. 2020); and elucidating that anti-discrimination policies increase the occupational attainment of all immigrant women but only highly skilled, non-Muslim immigrant men (Platt et al. 2022).

The connection between anti-discrimination policies and immigrant occupational attainment also points to a possible explanation of why immigrant–native intermarriage might be especially susceptible to societal attitudes. Negative attitudes toward outsiders often are a signal of deeper fissures within a society (Heizmann 2015; Rustenbach 2010). In the absence of strong policies protecting their interests, other means of protecting or promoting the well-being of foreign-born individuals—such as intermarriage—take on outsized importance (Song 2009).

Country and Individual-Level Hypotheses

Past research shows that immigrant–native intermarriage is associated with better employment outcomes, such as employment probability, labor market participation, and income (Bean and Stevens 2003; Stevens et al. 2012). However, occupational status remains underutilized when quantifying the effect of intermarriage on foreign-born workers’ labor force outcomes, despite its greater stability as an outcome measure (Ganzeboom and Treiman 2003) and providing better information on economic well-being and quality of working conditions.

Drawing on the literature reviewed in the previous section, we identify several possible mechanisms connecting immigrant–native intermarriage with immigrants’ occupational status. Native-born spouses can help immigrants acquire host country language skills and become more familiar with local culture, which strengthens their labor market position (Adserà and Pytlitkova 2015). Furthermore, intermarriage provides immigrants access to local social networks and insider information about high-quality employment opportunities (Furtado and Theodoropoulos 2010). Lastly, having a native-born spouse may signal a high degree of integration to employers, thus reducing barriers to harder-to-access segments of the labor market (Basu 2015; Ivory 2023). Therefore, we expect that:

• Hypothesis 1: Intermarried immigrants will have higher occupational statuses than those married to other immigrants.

The receiving context—most principally integration policies and societal attitudes—influences the positive relationship between immigrant–native intermarriage and an immigrant's occupational status. Existing studies suggest that the context of reception can either strengthen or weaken the association between intermarriage and occupational status (Rodríguez-García et al. 2015). More favorable integration policies reduce obstacles immigrants encounter in host country labor markets (Guzi et al. 2021). Labor market integration policies support the transferability of foreign credentials and, thus, enable immigrants to obtain jobs that match their skill set (Elo et al. forthcoming). Better citizenship policies open up more occupational opportunities that have historically excluded foreign nationals, such as public sector employment (Bratsberg et al. 2002), and anti-discrimination policies lessen immigrants’ disadvantages in the process of seeking employment (d'Appollonia 2008). The effects of supportive integration policies coincide with the labor market benefits associated with immigrant–native marriage, such as skill acquisition and local networks. In other words, immigrants in countries with highly inclusive integration policies would have less to benefit from marriages with natives. Therefore, we expect that:

• Hypothesis 2a: The positive association between intermarriage and occupational status will be weaker in countries with more favorable immigrant integration policies.

Conversely, better integration policies might strengthen the positive association between immigrant–native intermarriage and occupational status. Restrictive integration policies close immigrants’ access to primary labor sectors (d'Appollonia 2008) and are associated with increased occupational segregation (Harrison and Lloyd 2013). If immigrants are confined to marginal employment opportunities, intermarriage is likely to have little impact, whereas more favorable integration policies remove these constraints and maximize the benefits associated with intermarriage (Guzi et al. 2021). For example, immigrants without higher educational credentials from the receiving country may be excluded from occupations requiring postsecondary degrees, regardless of any sending-country credentials or transferable skills sharpened through marriage to a native-born individual (Adsera and Ferrer 2016). Thus, we propose a competing hypothesis:

• Hypothesis 2b: The positive association between intermarriage and occupational status will be stronger in countries with more favorable immigrant integration policies.

Positive societal attitudes toward immigrants—such as support for robust integration policies—are related to more favorable labor market conditions for immigrants (Fussell 2014). Foreign-born workers tend to experience less employment discrimination in societies with more welcoming attitudes toward immigrants (Bean et al. 2011). With less overt discrimination, immigrants should have an easier time building relationships with native-born individuals and learning useful labor market skills that enable them to access better quality occupations. In both cases, the relevance of immigrant–native intermarriage for occupational advancement diminishes. For immigrants living in more closed societies, intermarriage would provide considerable opportunities for enhancing occupational status. Most notably, personal connections and referrals obtained through native spouses’ social networks are one of the most effective routes for migrants to overcome discrimination and occupational segregation in societies that are more hostile toward outsiders (Ivory 2023). Following this logic, we hypothesize that:

• Hypothesis 3a: The positive association between intermarriage and occupational status will be weaker in countries with more inclusive societal attitudes toward immigrants.

In societies with more hostility toward outsiders, immigrant occupational attainment is more likely to be affected by prejudice from employers and other members of the host countries. Overt hostility prevents immigrants from accessing certain sectors of the labor market and undermines the utility of language skills, familiarity with the culture, and local networks in career advancement (Harrison and Lloyd 2013). Therefore, various advantages associated with immigrant–native marriage diminish when discrimination or outright exclusion of immigrants is pervasive within labor markets (Alba and Foner 2014), whereas these advantages play a more important role in employment seeking when prejudice against outsiders is low. These conditions could lead to worse spousal human capital and labor market–related resources, resulting in diminished intermarriage premiums. If the foregoing holds true, we would then expect a competing hypothesis to Hypothesis 3a:

• Hypothesis 3b: The positive association between intermarriage and occupational status will be stronger in countries with more inclusive societal attitudes toward immigrants.

Finally, contemporary research has demonstrated a heterogeneous and complex relationship between formal integration policies and immigrant-related attitudes (Callens and Meuleman 2017; Green et al. 2020). We expect integration policies and societal attitudes to have a complementary effect, as they are two distinct yet interconnected aspects of the context of reception. Immigrant-related attitudes should be more salient in countries with relatively weak formal integration policies and less influential in countries with strong institutional protections to promote inclusion and deter discrimination. Conversely, integration policies would matter most in the least immigrant-friendly societies where hostility toward immigrants remains high. Therefore, we offer the following hypotheses:

• Hypothesis 4: Societal attitudes moderate the positive association between immigrant–native intermarriage and occupational status in contexts with unfavorable integration policies.

• Hypothesis 5: Integration policies moderate the positive association between intermarriage and occupational status in contexts with hostile societal attitudes.

Data and Methods

Data

We used individual-level, cross-sectional microdata from EU-LFS (Eurostat 2018) and country-level contextual data from MIPEX (Solano and Huddleston 2020), ESS (2020), and the Eurostat database to test our hypotheses. EU-LFS is a collection of representative household surveys harmonized by Eurostat, the official statistical agency of the EU (Eurostat 2018). The EU-LFS collects quarterly data on employment status and sociodemographic characteristics of a representative sample of individuals aged 15 or older living in private households. Its large sample size makes it particularly useful to ensure an adequate representation of the foreign-born workers in our analysis.

We restricted our sample to partnered immigrants aged 20–64 who reported their occupation. Individuals were defined as partnered if their spouse could be identified within the same household; thus, we included both married and cohabiting couples, as cohabitation often confers the same rights and benefits as marriage in many EU member countries, and many cohabiting couples never marry yet share the same household resources as married couples. Immigrants were defined as third country nationals who were born in non-EU member states to exclude individuals with EU citizenship, as the latter have more guaranteed rights and access to the labor market than third country nationals.⁴ The EU-LFS does not provide sufficient information about the timing of union formation (i.e., whether it started in the host country). Therefore, we took an alternative approach to implement this crucial check by restricting our sample to immigrants who migrated by age 24, assuming that most individuals are unlikely to be married before 25 (see Basu 2015; Furtado and Theodoropoulos 2010). This step helped us to restrict our sample to immigrants who married after migrating to ensure they had the opportunity to select a native-born partner. Furthermore, we used alternative samples without the restriction of migration age and with individuals who migrated before the age of 19 and 29; the results were consistent and can be found in Table A1 in the online appendix (along with all other tables and figures designated with an “A”).

We used pooled observations from 11 waves of EU-LFS from 2008 to 2018. Separate models were run for males and females. As a sensitivity check, we also ran baseline models excluding country-years with fewer than 50 observations for either male or female subsamples and the results remained stable (see Table A2). After matching the available country-level indicators, the final sample consisted of 20 countries,⁵ 208 country-year combinations, and 425,436 individuals.

Dependent Variable

Our dependent variable, occupational status, was operationalized by the ISEI, which is constructed as “a weighted sum of mean education and mean income for each occupational group” (Ganzeboom et al. 1992:12). We recoded the three-digit International Standard Classification of Occupations (ISCO) ISCO-88 (or ISCO-08) available in the dataset into the ISEI and further divided it by 100 to facilitate model convergence. Thus, occupational status was a continuous variable ranging from .10 to .89, with higher values indicating occupations with higher socioeconomic statuses. The EU-LFS includes both the respondent's current occupation and previous occupation (if any) in the past eight years. We combined both variables to construct occupational status by using the previous occupation where current occupation was missing (for similar practices, see Heisig et al. 2018; Schmidt et al. 2022). Therefore, our dependent variable was less affected by the selection into employment. One potential problem with this approach is that immigrants who arrived less than eight years before enumeration may have reported their previous occupations in their country of origin. Additional checks revealed 1,139 respondents (0.27% of the sample) whose previous occupation referred to their countries of origin; excluding this population did not affect the findings.

Individual-Level Variables

Our main independent variable was spousal nativity, coded 1 if the respondent had a native-born spouse and 0 if their spouse was born in a non-EU member state. Several sociodemographic variables were included as controls following the literature convention. Age was provided by Eurostat in five-year bands to preserve anonymity. We included both age and its square term, age squared, as continuous variables. Education measured the highest level of education attained and was collapsed into three categories by Eurostat: lower secondary (ISCED 0–2), upper secondary (ISCED 3–4), and tertiary and above (ISCED 5–6). We constructed the categorical variable year of residence to measure the number of years the foreign-born respondent had resided in the host country (<3, 3–6, 7–10, 11–14, 15–19, and >19). Number of children younger than six years old was a continuous variable where higher values were associated with lower labor market participation for women. Finally, we included the region of birth to broadly control for the country of origin, because no more fine-grained information on the country of origin was available in the dataset to preserve anonymity.⁶

Country-Level Variables

Our macro-level indicators included commonly used measures of integration policy and immigrant-related attitudes, as well as seven control variables assessing labor market conditions and immigrant selection. Immigrant integration policy was measured by the MIPEX, a country-level index that evaluates the inclusiveness of countries’ integration policies for immigrants utilizing several distinct indicators. We used the average score of four domains most relevant to our study: access to nationality, anti-discrimination legislation/protections, labor market mobility, and permanent residency provisions.⁷Table A3 shows examples of the indicators used to construct the MIPEX variable. We rescaled this index to range from 0 (most unfavorable policy) to 1 (most favorable policy). Attitude toward immigrants was an indicator constructed using three questions from the ESS: (1) “Would you say it is generally bad or good for [country]'s economy that people come to live here from other countries?”; (2) “Would you say that [country]'s cultural life is generally undermined or enriched by people coming to live here from other countries?”; and (3) “‘Is [country] made a worse or a better place to live by people coming to live here from other countries?” After excluding foreign-born respondents, we averaged the responses to create a single, country-level indicator for each country in different years to capture the overall societal attitude toward immigrants. We rescaled this indicator to range from 0 to 1, with higher values representing more inclusive attitudes toward immigrants. Since the ESS collects data only in even years, the missing values were imputed on the basis of the average of data from nearby years.

Finally, we included a series of time-variant, country-level control variables that could be confounded with integration policy and immigrant attitude. Social protection expenditure, computed as the spending on social protection benefits per person (in euros), measured the expansiveness of the welfare state (Eurostat 2021); we rescaled this variable by dividing it by 1,000 to facilitate a more logical interpretation. Log of GDP per capita measured the overall economic development level (Eurostat 2020a). Employment rate estimated the labor market condition in the host country (Eurostat 2020b). Gini coefficient captured the level of income inequality. Unionization density controlled for the trade union's bargaining power (Visser 2019). Proportion of immigrants was defined as the percentage of the population born outside of the European Union. Lastly, the gender inequality index, developed by the United Nations Development Programme, ranged from 0 (complete gender equality) to 1 (complete gender inequality) (European Institute for Gender Equality 2020).

Given the hierarchical structure of our dataset, we estimated three-level linear regression models. Individuals (level-1 units) were nested within country-year combinations (level-2 units), which were further nested within countries (level-3 units). Such specification helped mitigate the problem of small-N estimations at the higher level, which is common among cross-national studies because the country-year combination yields a larger number of observations at the second level (Giesselmann and Schmidt-Catran 2019). We included a random slope for our key independent variable immigrant–native intermarriage at the country-year level. Separate models were run for male and female respondents. In both cases, our model took the following form:

where $yitj$ represents an individual i's occupational status in country-year t in country j. The vector $Xitj$ captures the individual-level explanatory variables. Year fixed effects were also included (⁠$γt$⁠). We grand-mean-centered the continuous individual-level explanatory variable (age, age squared, and number of children younger than six) by subtracting the full sample means from the raw scores. The random part of the model was captured by the three error terms: $vj$ (country level), $utj$ (country-year level), and $eitj$ (individual level).

Following the suggestion of Fairbrother (2014), we estimated a hybrid model (see also Bell et al. 2019; Schmidt-Catran 2016). In the analysis using pooled cross-national data, it is difficult to determine whether the effect of a country-level variable on the outcome variable is caused primarily by the differences between countries or by the difference over time within the same country. The hybrid model allowed us to separate the effect of country-level variables into both the between-country effect and the within-country effect while controlling for the compositional effect at the individual level.

In practice, we decomposed each country-level variable that changes over time (i.e., log of GDP per capita) into two distinct variables. We first calculated the average of each country-level variable across all relevant years—that is, the countrywide means that remain invariant over time (⁠$Z¯j$⁠). Next, we subtracted the countrywide means from the original country-level variables (⁠$Ztj−Z¯j$⁠). The variable $Z¯j$ varies among countries but remains stable over time within the same country, with its corresponding coefficient $β3$ capturing the between-country effect. The coefficient $β4$ for $Ztj−Z¯j$ represents the within-country effect—in other words, the association of change over time within the same country. By including both $Z¯j$ and $Ztj−Z¯j$ in the same model, we could simultaneously test the between and within effect of the country-level variables while not assuming they are the same effect. We followed the procedure described above to decompose all country-level variables. Finally, we further introduced cross-level interactions between spousal nativity, integration policy, and societal attitude to examine the moderating effects of receiving contexts.

Results

Table 1 presents descriptive characteristics of male and female immigrants by immigrant–native marital status. Intermarried immigrants of both genders have a higher average ISEI score than immigrants married to nonnatives—with higher scores associated with more prestigious jobs—and the difference is larger for women than for men. Intermarried immigrants are older, more educated, and have longer residency in the host country than those with a nonnative spouse; the spouses of intermarried immigrants also tend to be better educated and have a higher likelihood of being employed.

Table 2 shows substantial variations in most country-level variables. In addition, Figure 1 plots the changes in MIPEX—the higher the value, the stronger the national protections for immigrants—and attitude toward immigrants over time. Most countries experience larger variations in the attitude toward immigrants over time, compared with MIPEX. This is not surprising considering that integration policies change gradually. Another relevant observation is that these two variables do not always align closely. The correlation coefficient between MIPEX and attitude toward immigrants is .229, indicating a weak correlation. For example, Poland scores high on the attitude toward immigrants (.55) but has a relatively low MIPEX (.48). This pattern suggests that integration policies and societal attitudes capture distinct characteristics of the receiving context.

Table 3 provides results from the baseline multilevel models. Models 1 and 3 show the regression results without individual-level controls. On average across countries, immigrant–native intermarriage is associated with a 5.2-point increase (p < .001) in the ISEI for male immigrants and a 6.7-point increase (p < .001) for female immigrants. (To provide a reference point, the ISEI for accountants is 47 and that for secretaries is 42.) In Models 2 and 4, after controlling for age, number of children younger than six, education, year of residence, and region of origin, the coefficients for intermarriage diminish but remain positive and statistically significant: a 2.5-point increase (p < .001) in ISEI for males, and a 3.0-point increase (p < .001) for females. This demonstrates that intermarried immigrants have higher occupational statuses than those married to nonnatives, thus confirming Hypothesis 1, which posits a positive association between intermarriage and occupational status. The coefficients of the control variables align with our expectations. Being male, higher educational attainment, and longer residence periods are all associated with higher occupational status. Lastly, a positive attitude toward immigrants (between countries) is associated with higher occupational attainment, but only among females. The other measures of the receiving context did not have a statistically significant association.

The results from Table 3 mask the variations in the associations between immigrant–native intermarriage and occupational status across countries. Figure A1 reports results from country-specific linear regressions, showing substantial variations across countries in terms of both the coefficient size and the significance. In the next section, we use cross-level interactions to further investigate these trends.

After demonstrating the positive association between immigrant–native intermarriage and occupational status across the country-years, we now turn our attention to the moderation of the receiving context in terms of the policy and attitude environment. Table 4 presents the estimates of cross-level interactions between spousal nativity and the two measures of receiving context, using the same individual-level variables as in Table 3. As explained in the Data and Methods section, we included both the between effect and the within effect for each of the country-level variables. The models in Table 4 introduce only interactions between spousal nativity and the between-effect component of each country-level variable to better test our remaining hypotheses.

The positive and statistically significant coefficients of the interaction terms in Models 1 and 2 suggest that the association between immigrant–native intermarriage and occupational status is stronger in countries with more favorable migration integration policies, as measured by the MIPEX. The coefficient of the interaction term is larger for the female sample than for the male sample. Thus, our findings provide support for the positive relationship between the favorable immigration policy and the intermarriage premium (Hypothesis 2a) and not the converse (Hypothesis 2b). Overemphasizing the point estimates of interaction terms can obscure the substantial nature of the relationships (Brambor et al. 2006). Therefore, in Figures 2 and 3 we visualize the interaction by plotting the marginal effects of immigrant–native intermarriage over the ranges of values for our two country-level variables.

Consistent with the results in Table 4, in Figure 2 we observe positive slopes in the figures for integration policies. In countries with the most favorable policies (MIPEX of 87.69 for Portugal), female immigrants who married natives have a 5.2-point-higher ISEI, on average, than those who married other immigrants. In countries with the least favorable integration policies (MIPEX of 33.29 for Poland), immigrant–native intermarriage is associated with only a 0.9-point increase in ISEI. For the male sample, the association between intermarriage and occupational status is also substantially larger in countries with the highest MIPEX (4.2 points) than in those with the lowest (1.1 points).

The graphs for attitudes toward immigrants, on the other hand, do not depict a clear pattern. Figure 3 shows a slightly positive slope and a negative slope for the male and female samples, respectively. In both cases, the confidence intervals for the marginal effect largely overlap, suggesting insignificant moderation. The societal attitude toward immigrants does not seem to moderate the association between intermarriage and occupation status, as indicated by the insignificant interaction terms in Models 3 and 4 (Table 4). As a result, we find no evidence to support that inclusive social attitudes strengthen (Hypothesis 3a) or weaken (Hypothesis 3b) the intermarriage premium.

Finally, we present three-way interactions between immigrant–native intermarriage, integration policy, and immigration attitude to assess the combined influence of country-level factors on the labor market utility of intermarriage. The coefficients of the interaction terms in Models 5 and 6 in Table 4 are statistically significant for both male and female samples. To facilitate interpretation, Figures 4 and 5 illustrate the marginal effects of intermarriage as integration policy and immigration attitude move from minimum to maximum.

Figure 4 shows divergent moderating effects of immigration attitude in the context of the least, average, and most favorable integration policy. For the female sample, in the condition of the lowest MIPEX (.33), we observe a steep positive slope. The marginal effect of immigrant–native intermarriage increases from −0.05 to 0.03 as immigration attitude moves from the least to the most accepting. In other words, when favorable integration policies are absent, intermarriage is associated with a 5-point-lower ISEI in societies most hostile to immigrants and a 3-point-higher ISEI in societies most friendly to immigrants. The slopes for the interaction are slightly negative for average MIPEX and steeply negative for maximum MIPEX. Contrary to our expectation of a null finding, immigration attitude negatively moderates the association of intermarriage in the context of the most favorable integration policy: intermarriage is associated with an 18-point-higher ISEI in the most hostile immigrant attitude context and only a 1-point-higher ISEI in the friendliest context. Overall, the findings support our hypothesis that the interaction between immigration attitude and spousal nativity depends on integration policy conditions. The divergent slopes may also account for the lack of significance between immigration attitude and intermarriage, as sentiments may matter only when formal institutional protection is weak.

In Figure 5, we turn to the moderation of integration policy on the association between immigrant–native intermarriage and occupational attainment, holding immigration attitude at different levels. For both genders, the slope changes from positive to negative as we move from the least to the most accepting attitude environment. In conditions with maximum hostility toward immigrants, there is a positive interaction between integration policy and intermarriage. In the male sample, intermarriage is associated with a 5.2-point-lower ISEI in countries with the least favorable integration policies and a 14.3-point-higher ISEI in countries with the most favorable integration policies. However, the interaction turns negative and insignificant under conditions of minimum hostility: intermarriage is positively associated with occupational status (with an ISEI that is 0 to 4.3 points higher), but the association does not vary significantly depending on the integration policies. Therefore, integration policy matters most for intermarriage in places where immigrants are least welcomed, supporting Hypothesis 5—that integration policies moderate the positive association between intermarriage and occupational status in contexts with hostile societal attitudes.

We conducted a sensitivity test by dropping one country at a time. The results were consistent, except for the three-way interaction in the female sample when observations from Portugal are dropped. The coefficient remains negative but is statistically insignificant. Given these findings, we conclude that the three-way interaction between spousal nativity, integration policies, and societal attitudes holds under most circumstances.

We also conducted several additional analyses to ensure that our findings are robust to potential confounders and hold across subgroups. First, we used data from the EU-LFS ad hoc modules of 2008 and 2014 that ask supplemental questions about migrants and their immediate descendants. We then added two control variables that are potentially correlated with the unobserved characteristics: (1) self-rated host country language skills and (2) reasons for migrating (employment, study, asylum, family reasons, and other). These two additional controls do not change the results (see Table A4). Second, though we included an extensive list of relevant country-level controls, another potential confounder could be the host country's demand for skills. In other words, countries with stronger demands for skilled workers could simultaneously be more friendly to immigrants and have more favorable integration policies. Therefore, we further controlled for the country-level ISEI score and the difference between native-born and foreign-born ISEI scores. The results remain robust (see Table A5). Lastly, to account for the impacts of household features, we included additional controls for spousal education and spousal employment characteristics (employment status and occupational status) and find consistent results (see Table A6).

We also implemented several tests to examine the robustness of the cross-level interactions. First, multilevel models with a low number of top-level clusters (in our case, 20 countries) are at risk of influential outliers. Therefore, we reestimated the models with each country excluded one at a time. Figure A2 shows that the results for the two-way interactions are robust to the delete-one approach. The three-way interactions are robust, except for the Portuguese female sample, where the coefficient is still negative and becomes insignificant at the .05 level. This suggests that our findings on the interactions between spousal nativity, integration policies, and societal attitudes are sensitive to the sample selection and, therefore, should be interpreted with caution. Second, in the main analysis, countries with missing data on either MIPEX or attitude toward immigrants were excluded to maintain a consistent sample. We relaxed this restriction to expand the sample of countries: for example, countries with missing data on MIPEX were included in models analyzing cross-level interactions between spousal nativity and attitude toward immigrants. This approach increased the samples of countries slightly for models on MIPEX (22) and attitude toward immigrants (22); Table A7 shows that the results are consistent. Lastly, for immigrants from the pre-2000 period, their occupational transition may have occurred before the observation window in this analysis. To mitigate this concern, we excluded immigrants who started working for their current employers before 2000 and find consistent results (see Table A8).

Discussion

This study examines the relationship between immigrant–native intermarriage and foreign-born individuals’ occupational status and how this relationship is shaped by the context of reception in the receiving society via integration policies and societal attitudes toward foreigners. We contribute to the existing literature by drawing attention to moderating factors within the receiving context. Our multilevel analysis of 20 European countries establishes a positive association between immigrant–native intermarriage and foreign-born occupational status. More importantly, we demonstrate that this association is stronger in societies that have more favorable immigrant integration policies, and this result holds for both women and men within the sample. On the other hand, this association is not significantly affected by societal attitudes toward immigrants. Collectively, these findings help to unpack how the labor market benefit of intermarriage is amplified or diminished by the context of reception within receiving societies. These findings will be of particular use to migration, marriage, and gender scholars investigating the nature of intersectional inequality and factors that most effectively moderate its severity.

We recognize the possibility of selectivity bias with intermarried immigrants and, thus, focus on only the association between immigrant–native intermarriage and employment outcomes and refrain from making any causal claims. Selectivity bias may exist at the country level, as well; immigrants could be incentivized to move to countries with more favorable integration policies and leave ones with exclusive policies. While we cannot conclusively rule out selectivity as an issue, we have reason to believe that it has a limited effect on our results. Most countries in our sample have relatively restrictive entry policies for non-EU nationals. Therefore, immigrants cannot freely move between countries once they enter. This also means that they cannot simply choose among European countries on the basis of integration policies. Instead, the choice of destination countries is likely based on a variety of factors, such as entry requirements, availability of resources, and overseas social networks (Castles 2007; Haug 2008; Paul 2017).

Societal attitude toward immigrants is an important indicator of the receiving context but, unlike integration policies, does not play a significant role in shaping the association between immigrant–native intermarriage and immigrant occupational status. This suggests that policy conditions might have more relevance than societal attitudes in the context of intermarriage. Integration policies can have direct impacts on immigrants’ employment outcomes by increasing or reducing access to better occupations (Kogan 2016). The influence of societal attitudes is indirect and complex. The overall openness to foreigners may not translate directly to individual circumstances, and a country-level measure of openness also masks the heterogeneity at the local level, which should be explored in future research using more fine-grained measurement of contextual attitude environments.

Some limitations of this study include that we were unable to assess causal effects of intermarriage on occupational status and we cannot exclude the possibility of reverse causality. For example, immigrants who are employed in jobs of high occupational status may have a higher likelihood of encountering native-born individuals, given that immigrants are overrepresented in high- and low-status occupations (Kalmijn 1998). We address this issue by controlling for the proportion of opposite-sex immigrants in each occupation (see Table A9). We recognize that our data and research design are insufficient to allow us to make causal arguments and, thus, have been careful to present our findings in terms of associations. Even though we cannot make causal claims, the association between immigrant–native intermarriage and occupational status is quite robust and holds under every treatment within the analysis.

Future research should also investigate the mechanisms behind the association between integration policies and labor market outcomes. We hypothesize that favorable integration policies aid immigrants by opening access to a greater range of job opportunities and, therefore, maximize the advantages associated with intermarriage. Because of data limitations, we did not test these mechanisms directly. Information about where respondents received their education credentials, public versus private sector employment, and reasons for migration can help address the precise mechanism. Furthermore, the hybrid models in this study tested “between effects” rather than “within effects.” In other words, we focused on the variation between countries instead of the change over time within the same countries. This is largely because there are a limited number of variations in integration policies within most countries in the time frame of our study, which is further limited by the availability of the MIPEX database. This should be a promising direction for future research when new data become available.

There is still much research to be done to further map out how receiving societies augment the labor market benefits of immigrant–native intermarriage for foreign-born workers. First and foremost is the addition of more fine-grained data on the ethnicity of foreign-born individuals within the labor market. Collecting sensitive information at such a high resolution may inadvertently identify individuals in particular places who had been promised anonymity by the data collectors. This may make them vulnerable to anti-immigrant threats. However, safeguards for preventing that from happening are well-established in the United States, where highly detailed ethnoracial data are collected and made available to researchers. The next substantive area of concern involves assessing immigrant–native intermarriage and its labor market benefits across a greater variety of countries. Excellent data sources, such as the international version of IPUMS, are fast making the ability to do this a reality. Until we move beyond countries in North America and Europe, it will be impossible to determine whether the underlying theories supporting why immigrant–native intermarriage results in labor market benefits are truly universal or more of an artifact of the specific nature of interpersonal relations in Western societies.

Acknowledgments

The authors would like to express deep gratitude to both the anonymous reviewers and the Demography editorial team for their generous and helpful comments. Substantive feedback was also provided by Guilherme Chihaya Da Silva, Hillary Holbrow, Vida Maralani, C. Matthew Snipp, and Kelly Musick. We would like to thank Stephen Parry and May Boggess from the Cornell Statistical Consulting Unit and Jonathan Bohan from the Cornell Restricted Access Data Center for technical assistance with the analysis. Last, but certainly not least, this research would not have been possible without access to the European Union Labor Force Survey, and we thank Eurostat for their continued guidance on this project. Authorship is shared equally between the coauthors, as are all errors or omissions.

Notes

References